Impact of HSDPA Implementation on UMTS Interference and Cell Coverage. Mónica Rute Cardoso Antunes

Impact of HSDPA Implementation on UMTS Interference and Cell Coverage Mónica Rute Cardoso Antunes Dissertation submitted for obtaining the degree of Master in Electrical and Computer Engineering Jury Supervisor: President: Members: Prof. Luís M. Correia Prof. António Rodrigues Prof. António Topa Eng. Carlos Caseiro December 2007

i To Rui

ii

Acknowledgements First of all, I would like to thank Prof. Luís Correia for giving me this opportunity to share his knowledge and learn from it. I thank him not only for his singular professional orientation, but also for his constant help, advice and sharing of knowledge, which became a reference along this period. And the dedicated and disciplined supervision of this thesis always kept me motivated to walk forward. To Vodafone, especially to Carlos Caseiro, Marco Serrazina and Pedro Lourenço, for their useful suggestions throughout the development of this thesis, for their time and patience, and for giving me a more realistic vision of the technology. To all GROW members, for all the constructive critics, for always being available and willing to help, for showing how useful it is to work as a team, sharing experiences and knowledge. Also, to have contact with other areas of mobile and wireless communications. A special thanks to Daniel Sebastião, for the patient help with OPNET, and for always having a positive word in the right moments. To Pedro Sobral, João Lopes, and Luís Salvado for all their precious help, for the important talks, useful advices and necessary discussions to put in order all the thoughts and ideas, for the great company in difficult moments, for their friendship. I also would like to thank my friends and family for all the essential support, strength, hope, caring and understanding. It was their guidance along this path, their support each day that made this work possible. Mónica Antunes iii

iv

Abstract The goal of this thesis was to analyse how UMTS behaves when HSDPA is implemented. So, a theoretical model was developed calculating network capacity, for a sector of a cell, focusing on the analysis of parameters, such as number of served users, throughput, and cell radius. Two scenarios were assessed with different users distribution. With OPNET Modeler, Release 99 was assessed by increasing the number of users, and for two different scenarios that were distinguished by the weight assigned to each service. For the theoretical model, two analyses are done: one until the network saturates, and other for a percentage of rejected users of 5%. For the latter, one verifies that the limiting factor is the spreading factor. An analysis of the global traffic processed by the Node B was also made, obtaining 350 MB/h as a maximum for a heavy data profile, and 140 MB/h for a light one, for a service profile based on voice. In OPNET Modeler, a single-sector cell was evaluated for an increasing number of users, and for two different service scenarios. Global and application related parameters were assessed. As OPNET gives priority to Conversational and Streaming classes, above 15 users, the percentage of queued requests is higher than 80%, the remaining being granted. Only Interactive and Background classes contribute for the percentage of queued requests. As expected, the application throughput is low, and response times are high, not only for data services but also for voice ones, thus, considering VoIP not being recommendable in Release 99. UMTS FDD, HSDPA, OPNET Modeler, Multiservice Traffic, Capacity v

Resumo O objectivo desta tese foi a análise do comportamento da rede UMTS com a implementação de HSDPA. Um modelo teórico do cálculo da capacidade foi desenvolvido, analisando apenas um sector da célula, focando-se na análise de parâmetros como o número de utilizadores servidos pelo Node B, o ritmo de transmissão e o raio da célula. Foram definidos dois cenários com diferentes disposições de utilizadores. Recorrendo ao simulador OPNET Modeler, a Release 99 foi analisada variando o número de utilizadores, e definindo dois cenários que se diferenciavam entre si pelo peso atribuído a cada serviço. No modelo teórico, a análise foi feita de duas formas: até alcançar a saturação da rede e até um grau de serviço de 5%. Para o último caso, verificou-se que o factor mais limitativo é o factor de espalhamento. Foi ainda efectuada uma análise do tráfego global, obtendo-se no máximo 350 MB/h para um perfil centrado em dados e cerca de 140 MB/h para um perfil de serviços centrado na voz. Foi simulada uma célula com um sector. Foram analisados resultados em termos globais e por aplicação. Dando o OPNET prioridade às classes Conversational e Streaming, a partir de 15 utilizadores, a percentagem de pedidos em espera é superior a 80% sendo que os restantes pedidos foram admitidos. O ritmo de transmissão para serviços de dados é baixo, e o tempo de espera é elevado, este último não só para os dados mas também para a voz, tornando VoIP uma tecnologia pouco recomendável sobre Release 99. UMTS FDD, HSDPA, OPNET, Tráfego Multiserviço, Capacidade vi

Table of Contents Acknowledgements...iii Abstract...v Resumo...vi Table of Contents...vii List of Tables...xii List of Acronyms...xiv List of Symbols...xvii List of Programmes...xx 1 Introduction...1 2 UMTS Basic Aspects...7 2.1 The system...8 2.1.1 Network Architecture...8 2.1.2 Radio interface...9 2.1.3 Capacity and Interference...11 2.2 Services and Applications...13 2.3 HSDPA...15 2.3.1 Channels...15 2.3.2 HSDPA Performance...17 2.3.3 Radio Resource Management...19 2.4 Performance Analysis...20 3 Capacity and Interference Model...23 3.1 Model...24 3.2 Scenarios...28 3.3 Analysis of results...31 3.3.1 User distribution...31 3.3.2 Release 99...34 3.3.3 HSDPA Shared Carrier...37 3.3.4 HSDPA Dedicated Carrier...40 3.3.5 Scenario 2...42 vii

3.4 Global Traffic...43 4 Simulator...47 4.1 OPNET Overview...48 4.2 UMTS...50 4.3 Input Parameters...54 4.4 Analysis of Results...56 4.4.1 Global Results...57 4.4.2 Applications...60 4.5 Comparison with Theoretical Model...65 5 Conclusions...69 Annex A Link Budget...75 Annex B Model...81 Annex C Capacity and Interference Model Results...85 Annex D Application and Profile Definitions...91 Annex E Variation of the Number of Users per Scenario...97 Annex F Parameter Definitions...99 References...101 O tamanho disto deve ser 11pt viii

List of Figures Figure 2.1. UMTS Network Architecture, [HoTo04]...8 Figure 2.2. Relation between spreading and scrambling, [Corr06]...10 Figure 2.3. Channels needed for HSDPA operation in Release 5, [HoTo06]...16 Figure 2.4. Single-user performance with 16QAM/QPSK and with QPSK-only, [HoTo06]....17 Figure 2.5. HSDPA data rate as a function of the average HS-DSCH SINR, [Pede05]...18 Figure 3.1. Scenarios for capacity calculation...28 Figure 3.2. User efficiency, as a function of the number of users...32 Figure 3.3. User distribution for Voice Centric profile and QoS priority policy...32 Figure 3.4. User distribution for Voice Centric profile and Proportional priority policy...33 Figure 3.5. Number of served users for different profiles...33 Figure 3.6. Average throughput for Release 99 users....34 Figure 3.7. Load factor for Voice Centric profile and Proportional priority policy...34 Figure 3.8. Load Factor for Voice Centric profile...35 Figure 3.9. Load Factor for Data Centric profile...35 Figure 3.10. Characteristics for Voice single service....36 Figure 3.11. Cell radius for Release 99...36 Figure 3.12. Cell radius for different environments in Release 99, for Voice Centric profile...37 Figure 3.13. Target and effective throughput for both priority policies, and for the Voice Centric profile...37 Figure 3.14. Throughput efficiency for Voice Centric profile....38 Figure 3.15. Maximum cell radius as a function of the number of users...38 Figure 3.16. Cell radius for different environments in HSDPA, for Voice Centric profile...39 Figure 3.17. Difference between the cell radius for pedestrian and indoor environment....39 Figure 3.18. User distribution for Data Centric profile and QoS priority policy...40 Figure 3.19. User distribution for Data Centric profile and Proportional priority policy...40 Figure 3.20. Target and effective throughput profile and for QoS priority policy...41 Figure 3.21. Target and effective throughput for Proportional priority policy...41 Figure 3.22. Maximum cell radius for Data Centric profile....41 Figure 3.23. Cell radius for different environments in HSDPA, for Data Centric profile....42 Figure 3.24. Percentage of rejected users...44 Figure 3.25. Global traffic processed by the Node B for Voice Centric profile...45 Figure 3.26. Global traffic processed by the Node B for Data Centric profile...45 Figure 4.1. Example of a scenario using the Project Editor....49 Figure 4.2. Example of UMTS workstation architecture in the Node Editor...50 Figure 4.3. Example of process model in the Process Editor...51 ix

Figure 4.4. Simple UMTS network using raw traffic generation...52 Figure 4.5. Simple UMTS network using application traffic...52 Figure 4.6. UMTS workstation node model...53 Figure 4.7. Node B node model...54 Figure 4.8. Example of the network architecture...55 Figure 4.9. Calls per hour for Data Centric profile and Light scenario...57 Figure 4.10. Percentage of requests granted and queued...57 Figure 4.11. Percentage of requests granted per QoS class...58 Figure 4.12. Percentage of requests queued per QoS class....58 Figure 4.13. Node B total DL throughput...59 Figure 4.14. Average number of DL transmissions required...59 Figure 4.15. Maximum number of DL transmissions required...60 Figure 4.16. Average throughput received in the MT, for VoIP applications...60 Figure 4.17. Average throughput received in the MT, for Streaming applications...61 Figure 4.18. Packet end-to-end delay, for VoIP applications...61 Figure 4.19. Packet end-to-end delay, for Streaming applications...61 Figure 4.20. Average throughput received in the MT, for FTP application...62 Figure 4.21. Average throughput received in the MT, for HTTP application...62 Figure 4.22. Download response time, for FTP application....63 Figure 4.23. Page response time, for HTTP application....63 Figure 4.24. Average throughput received in the MT, for E-mail application...64 Figure 4.25. Average throughput received in the MT, for MMS application...64 Figure 4.26. Download response time, for E-mail and MMS applications...65 Figure 4.27. Global traffic in the Node B as a function of the number of users, for Voice Centric profile...66 Figure 4.28. Global traffic in the Node B as a function of the number of users, for Data Centric profile...66 Figure 4.29. Reduction in the traffic comparing Heavy and Light scenarios...67 Figure A.1. Approximated SINR values for 10 and 15 HS-PDSCH codes...78 Figure A.2. Approximated throughput values for 10 and 15 HS-PDSCH codes...79 Figure C.1. User distribution for Data Centric profile, QoS priority policy and limit load factor 0.7....86 Figure C.2. User distribution for Data Centric profile, Proportional priority policy and limit load factor 0.7...86 Figure C.3. Target and effective throughput profile and for QoS priority policy, for the Data Centric profile...87 Figure C.4. Throughput efficiency for Data Centric profile...87 Figure C.5. User distribution for Voice Centric profile and QoS priority policy...87 Figure C.6. User distribution for Voice Centric profile and Proportional priority policy...88 Figure C.7. Throughput for Release 99, for the limitation of users rejected of 5%...88 Figure C.8. Load factor for Voice Centric profile, for the limitation of users rejected of 5%...88 Figure C.9. Load factor for Data Centric profile, for the limitation of users rejected of 5%...88 x

Figure C.10. Cell radius for Release 99, for the limitation of users rejected of 5%...89 Figure C.11. Target and effective throughput for QoS priority policy, for the Data Centric profile, and for the limitation of users rejected of 5%...89 Figure C.12. HSDPA cell radius for Data Centric profile, and for the limitation of users rejected of 5%...89 xi

List of Tables Table 2.1. Functionality of the channelisation and scrambling codes, [Corr06]...10 Table 2.2. QoS classes main parameters and characteristics, [3GPP06a]...14 Table 2.3. HSDPA terminal capability categories, [HoTo06]...16 Table 3.1. Spreading factor allocation priority table...25 Table 3.2. QoS priority classes...26 Table 3.3. Throughput per code and maximum number of HS-PDSCH codes...26 Table 3.4. Voice centric and data centric service profiles...29 Table 3.5. PS services target throughput, in HSDPA...29 Table 3.6. PS services target throughput....29 Table 3.7. Default values used in link budget...30 Table 3.8. Fading margins and indoor penetration values, [EsPe06]...30 Table 3.9. Eb N0 values for the different services considered in Release 99, [CoLa06]...30 Table 3.10. Activity factor values for the different services considered in Release 99...31 Table 3.11. Corresponding frequencies of each carrier used for simulations...31 Table 3.12. Maximum cell radius for Scenario 2...43 Table 3.13. Number of calls per hour for each scenario...44 Table 3.14. Mean call duration and mean file volumes...44 Table 4.1. Throughput limitation of each QoS class...55 Table 4.2. Application and profile main parameters...56 Table 4.3. Trend constants for OPNET results and theoretical model...66 Table A.1. Relative mean error and variance for SINR values...77 Table A.2. Relative mean error and variance for throughput values...79 Table B.1. Parameters used for propagation model...83 Table D.1. Voice application definitions....92 Table D.2. Streaming application definitions...92 Table D.3. FTP application definitions...93 Table D.4. HTTP application definitions...93 Table D.5. E-mail application definitions...93 Table D.6. MMS application definitions...94 Table D.7. VoIP user profile definition...94 Table D.8. Streaming user profile definition...94 Table D.9. FTP user profile definition...95 Table D.10. HTTP user profile definition...95 xii

Table D.11. E-mail user profile definition....95 Table D.12. MMS user profile definition...96 Table E.1. Effective number of calls per hour...98 xiii

List of Acronyms 3GPP AAL AICH AM AMC AMR ARP ARQ ATM BCH BS BWA CDMA CN CPCH CPICH CPU CQI CS DCH DES DL DSCH DSL DTX EDGE EIRP FACH FDD FSM GGSN GMM GMSC GPRS 3 rd Generation Partnership Project ATM Layer Adaptation Acquisition Indicator Channel Acknowledged Mode Adaptive Modulation and Coding Adaptive Multi Rate Allocation/Retention Priorities Automatic Repeat Request Asynchronous Transfer Mode Broadcast Channel Base Station Broadband Wireless Access Code Division Multiple Access Core Network Common Packet Channel Common Pilot Channel Central Processing Unit Channel Quality Indicator Circuit Switched Dedicated Channel Discrete Event Simulation Downlink Downlink Shared Channel Digital Subscriber Loop Discontinuous Transmission Enhanced Data rates for GSM Evolution Equivalent Isotropic Radiated Power Forward Access Channel Frequency Division Duplex Finite State Machine Gateway GPRS Support Node GPRS Mobility Management Gateway MSC General Packet Radio System xiv

GSM GUI HARQ HLR HSDPA HSOPA HSPA HSUPA HS-DPCCH HS-DSCH HS-PDSCH HS-SCCH IEEE IP ITU LTE MAC MAC-hs ME MIMO MMS MT MS MSC NBAP OFDM OPNET OVSF PCH PDCP PDP PRACH PS PSTN QAM QoS QPSK RAB RACH RAN Global System for Mobile Communications Graphical User Interface Hybrid Automatic Repeat Request Home Location Register High Speed Downlink Packet Access High Speed OFDM Packet Access High Speed Packet Access High Speed Uplink Packet Access High-Speed Dedicated Physical Control Channel High-Speed Downlink Shared Channel High-Speed Physical Downlink Shared Channel High-Speed Shared Control Channel Institute of Electrical and Electronics Engineers Internet Protocol International Telecommunication Union Long Term Evolution Medium Access Control Medium Access Control high speed Mobile Equipment Multiple Input Multiple Output Multimedia Message Service Mobile Terminal Mobile Station Mobile Services Switching Centre Node B Application Protocol Orthogonal Frequency Division Multiplexing Optimized Network Engineering Tool Orthogonal Variable Spreading Factor Paging Channel Packet Data Convergence Protocol Packet Data Protocol Physical Random Access Channel Packet Switched Public Switched Telephone Network Quadrature Amplitude Modulation Quality of Service Quadrature Phase-Shift Keying Radio Access Bearer Random Access Channel Radio Access Network xv

RIP RLC RNC RNS RRM RSVP SCCPCH SF SGSN SINR SIP SIR SMS SRB TCP TDD TPAL ToS TTI UDP UE UL UMTS USIM UTRAN VLR VoIP WCDMA WiBro Wi-Fi WiMAX Routing Information Protocol Radio Link Control Radio Network Controller Radio Network Sub-system Radio Resource Management Resource Reservation Protocol Secondary Common Control Physical Channel Spreading Factor Serving GPRS Support Node Signal-to-Interference-plus-Noise Ratio Session Initiation Protocol Signal-to-Interference Ratio Short Message Service Signalling Radio Bearer Transport Control Protocol Time Division Duplex Transport Adaptation Layer Type of Service Time Transmission Interval User Datagram Protocol User Equipment Uplink Universal Mobile Telecommunications System UMTS Subscriber Identity Module UMTS Terrestrial Radio Access Network Visitor Location Register Voice Over IP Wideband CDMA Wireless Broadband Wireless Fidelity Worldwide Interoperability for Microwave Access xvi

List of Symbols α DL orthogonality factor. α j Orthogonality factor for user j. η Load Factor. η DL Global load factor of a cell in DL. η HS HSDPA throughput efficiency. η N u Number of served users ratio. η UL Global load factor of a cell in UL. f Signal bandwidth. R Distance between users. λ Number of calls per hour. λ ef Effective calls per hour. λ i number of calls per hour for service i. ρ Signal-to-interference-plus-noise ratio θ C θ u ψ C u d E b f F Throughput per code. Throughput per user. Orientation angle. Number of HS-PDSCH codes per user. Distance between the cell radius for the pedestrian and indoor environments. Bit Energy Frequency. Noise figure. F a j Activity factor for user j. G P G r Processing gain. Gain of receiving antenna. G SHO Soft handover gain. G t Gain of transmitting antenna. xvii

H B h b h m i Node B height. Building height. MT height. DL Inter-to-intra cell interference ratio. I inter n Normalised inter-cell interference. I inter n j Normalised inter-cell interference of user j. L C Losses in cable between transmitter and antenna. L int Indoor penetration. L P Path loss. L p j Average cell path-loss. L P total Total path loss. L tm L tt L u Losses due to diffraction from the roof to the MT. Losses due to propagation over roof tops. Losses due to user. M FF Fast fading margin. M I Interference Margin. M SF Slow fading margin. N 0 Noise power spectral density. N au Number of active users per cell. N C Number of connections per cell. N i Number of served users for service i. HS N MAX Maximum number of HS-PDSCH codes. N RF Average noise power. N s N u Number of users of the corresponding service. Number of users. HS N u Number of HSDPA users. serv N u Number of served users. p HSDPA HSDPA power percentage. xviii

P r P Rx Power available at the receiving antenna. Power at the input of the receiver. P Rx min Receiver sensitivity. P t P Tx Power fed to transmitting antenna. Transmitter output power. BS P Tx Total transmission power. R Cell radius. ef R b Effective throughput. tg R b Target throughput. R b j Bit rate of user j. R c WCDMA chip rate. R max Maximum cell radius. T Global traffic processed by the Node B. V Mean file volume for service i. i w B w s Building separation. Street width. xix

List of Programmes MatLab Microsoft Office Excel Microsoft Office Word Microsoft Visual Studio OPNET Modeler xx

1 Introduction This chapter gives a brief overview of the work. Before establishing work targets and original contributions, scope and motivations are brought up. At the end of the chapter, the work structure is provided. 1

The development of the Mobile Communication Systems caused a revolution in the way people were used to communicate. One of the most important advantages that mobile systems brought over the fixed ones was mobility. Nowadays, people can always be reachable, anytime, anywhere. Mobile Communication Systems developed from the so-called First Generation systems to the ones used nowadays, Third Generation systems. First Generation systems were analogue and only allowed voice communications. With the arrival of the Second Generation systems, already digital ones, although the main interest of the users was still voice communication, the importance of other types of service has grown. Examples of these services can be short text messaging and enabling the access to data networks. As the interest of the public grows around multimedia services, new technologies arise, providing an access to these kinds of services with an improved quality, turning the telecommunications business a very competitive and lucrative one, which benefits both end users and providers. Figure 1.1 depicts the evolution of technologies and the trend in mobile wireless access, based on [UMFo07]. As an example of this spreading of mobile communications, according to ANACOM, the First Generation appeared in Portugal in 1989, and at the end of that year 2800 subscribers were counted. This number has grown constantly, and at the end of 2006 there were 12.2 million subscribers in Portugal, with a penetration rate of 115.7%, [ANAC07]. Figure 1.1. Evolution of mobile wireless access technologies, [UMFo07]. With this growing interest, networks supporting data communication with better quality started arising. With Global System for Mobile Communications (GSM), data services were supported, although the 2

throughputs were very low. So, General Packet Radio Service (GPRS) appeared bringing a new variety of multimedia services and enabling a better quality in the access to data networks. This quality improved with the appearance of Universal Mobile Telecommunication System (UMTS), which is the European standard for the Third Generation systems. UMTS Release 99 enabled a very good access to data networks, providing a maximum of 384 kbit/s data rate in downlink (DL), [HoTo04], which brought a new component to the mobile communications market, turning mobile operators also into Internet providers, competing with fixed operators not in terms of throughput but in terms of mobility. Figure 1.2 shows the Wideband Code Division Multiple Access (WCDMA) subscribers growth from January 2004 to February 2006. Figure 1.2. Number of WCDMA subscribers growth, [HoTo06]. With the increasing interest on data based services, High Speed Downlink Packet Access (HSDPA) was developed, its first specification being integrated as part of 3GPP Release 5, [HoTo06], and was launched on March 2002. When the first HSDPA networks became available at the end of 2005, data rates were able to reach 1.8 Mbit/s, increasing to 3.6 Mbit/s in 2006, and nowadays 7.2 Mbit/s are already available. According to the UMTS Forum, Portugal is among the top ten countries ranked by WCDMA penetration, with 25% WCDMA, and 1% HSPA, [UMFo07]. With these improvements in the DL channel, and, hence, the growing number of users using data based services, Release 99 has not enough capacity to answer the expectations for the uplink (UL) channel. So, in December 2004 a new specification was launched, Release 6, which characterised High Speed Uplink Packet Access (HSUPA), enabling data rates up to 2 Mbit/s in an initial phase, [HoTo06]. Some new features introduced with HSDPA, which enhances its performance, are the adaptive modulation and coding (AMC), adjusting the data rate concerning the available channel quality, the Hybrid Automatic Repeat Request (HARQ) mechanism, performing retransmission in the physical layer, and the medium access control layer (MAC-hs) located in the Node B. Also a new Transmission 3

Time Interval (TTI) of 2 ms is implemented, allowing the system to be more reactive to changes in the radio channel conditions, thus, performing a better allocation of resources among users. The main characteristic that enables this technology is the new supported modulation, 16-QAM, contrarily to Release 99, which only supported Quadrature Phase Shift Keying (QPSK). Only for HSDPA use, a new channel is implemented, which can be shared among various users, [HoTo06]. Concerning HSUPA, the most important introduced characteristics are the fast Node B scheduling, a new dedicated channel, and, similarly to HSDPA, the introduction of the HARQ mechanism. With the new enhanced characteristics brought by High Speed Packet Access (HSPA), new services arise, as Voice over IP (VoIP) or multiplayer gaming, and existing ones are improved, as peer-to-peer applications and streaming video. Figure 1.3 depicts the evolution enabled by HSPA to the services supported by this technology. Figure 1.3. Evolution in services with HSPA, [UMFo07]. The evolution in technology improvements is constant, so, combining Orthogonal Frequency Division Multiplexing (OFDM) and Multiple Input Multiple Output (MIMO) antennas, High Speed OFDM Packet Access (HSOPA) appeared, enabling throughputs even higher than HSPA and a significant decrease in the cost per bit, and should become an attractive alternative to providers and end-users, [Nort05]. Already considering a Fourth Generation technology, 3GPP is already working in Long Term Evolution (LTE) technology, which will enable increased peak data rates, with instantaneous maximums of 100 Mbit/s in DL and 50 Mbit/s in UL, both in a 20 MHz frequency band. Another advantage is the compatibility with earlier releases and with other systems, [3GPP07]. 4

The goal of this thesis is to analyse the impact of HSDPA implementation on the behaviour of UMTS networks, in terms of capacity and interference. For this, a model that calculates network capacity in terms of served users, throughput and cell radius was developed and implemented. The obtained results are compared with simulation results, which were obtained by using a modelling and simulating tool named OPNET Modeler. Only Release 99 was simulated, since OPNET does not have HSDPA implemented. This thesis is composed of 5 chapters, including this Introduction. Chapter 2 presents an overview of UMTS basic aspects, describing how the system works focusing on network architecture, the radio interface, and capacity and interference. Nowadays in use services and applications are presented. HSDPA is described, focusing on the new channels that had to be introduced, the system performance and on radio resource management. Finally, a performance analysis is made describing the state of the art. In Chapter 3 the model for theoretical calculations is presented. The chosen scenarios and correspondent parameters are described and an analysis of results is made, firstly analysing the users distribution, then Release 99 and HSDPA and finally, gathering the obtained results, analysing the global traffic of the network. Chapter 4 introduces the OPNET Modeler simulator, describing its functionalities focusing in the UMTS modules. Then the options made for simulations, as network architecture, profiles and applications parameters, and scenarios are presented. The analysis of results is made separating global results of the network and applications results. Finally, a comparison between the results of the simulator and the theoretical model is made. Finally in Chapter 5 some conclusions are drawn and suggestions of future work presented. 5

6

2 UMTS Basic Aspects This chapter provides an overview of UMTS basic aspects. In Section 2.1, the system is described, focussing on its architecture, aspects of the radio interfaces, capacity and interference. In Section 2.2, services and applications are presented. HSDPA is presented in Section 2.3, which describes the new channels, performance and radio resource management. Finally, in Section 2.4, the current state of the art is presented. 7

2.1 The system 2.1.1 Network Architecture The elements that compose the UMTS network can be divided into three groups, [HoTo04], as depicted in Figure 2.1., according to their functions: the User Equipment (UE), the UMTS Terrestrial Radio Access Network (UTRAN) and the Core Network (CN). Figure 2.1. UMTS Network Architecture, [HoTo04]. The UE is responsible for the radio communications between the user and the network, including two elements: the UMTS Subscriber Identify Module (USIM), which is a smartcard containing information about the subscriber and authentication operations, and the Mobile Equipment (ME), which is the radio terminal that makes the connection between UE and UTRAN possible, over the Uu Interface, i.e., the WCDMA radio interface. The USIM and the ME are connected to each other via an electrical interface, named Cu. UTRAN deals with the radio related functionalities, comprising one or more sub-systems, called Radio Network Sub-systems (RNS). Each RNS is composed of one Radio Network Controller (RNC) and one or several Node Bs, i.e., Base Stations (BSs). The RNC is responsible for the control of the radio resources, and the Node B processes the information flow between UE and RNC. UE and UTRAN protocols were designed for the needs of the WCDMA radio technology, but the UMTS CN is an adaptation of the GSM one. The CN consists of the following elements: Home Location Register (HLR): a database where the operator subscribers information is stored, consisting mainly of the users service profile and their location. 8

Mobile Service Switching Centre (MSC): dealing with the switching in the Circuit Switched (CS) domain. Visitor Location Register (VLR): database with information on active users. Gateway MSC (GMSC): linking the CN to CS external networks. Serving GPRS Support Node (SGSN): performing a function identical to MSC/VLR, but concerning Packet Switched (PS) services. Gateway GPRS Support Node (GGSN): linking the CN to PS external networks. The CN switches and routes calls or data connections to external networks, which can be of two types: CS networks, such as Public Switched Telephone Network (PSTN), or PS ones, such as the Internet. One of the main factors for the new design of the UE and UTRAN modules is that the latter was required to support soft handovers. In GSM, only hard handover occurs, in which the connection from a Mobile Terminal (MT) is reassigned from one BS to another without being connected simultaneously to both. In UMTS, hard handover can be either inter-frequency or inter-system: in the former the MT is transferred from one frequency carrier to another, which involves the use of several carriers per BS; in the latter, the MT is transferred from UMTS to another system, currently GSM, enabling load balancing and coverage enhancements. UMTS supports the other two types of handover: soft handover, where the MT is connected to more than one BS at the same time; softer handover, similar to soft handover, but only when the MT is transferred from one sector of a cell to another sector of the same cell. 2.1.2 Radio interface In UMTS, the air interface is based on WCDMA, which was specified by the 3 rd Generation Partnership Project (3GPP), [3GPP07]. There can be two modes of operation: Frequency Division Duplex (FDD) and Time Division Duplex (TDD). In this work, only the FDD mode is considered, since the TDD mode, which was designed to provide high data rates, was replaced by HSPA in FDD (which is addressed later). WCDMA is a wideband multiple access technique based on CDMA, [HoTo04], which leads to a very wide bandwidth, a carrier bandwidth of 4.4 MHz being achieved with a chip rate of 3.84 Mcps. Another important characteristic of WCDMA, mainly for PS data services, is the possibility of performing variable user data rates; the data rate is constant over each 10 ms frame, but it can be different from frame to frame. Spreading is used to separate the physical data and control channels in UL and to distinguish the connections to different users within one cell in DL, consisting of two operations: channelisation and 9